Long-Term Prostacyclin for Pulmonary Hypertension With Associated Congenital Heart Defects
Background—Although long-term prostacyclin (PGI2) has been shown to improve hemodynamics, quality of life, and survival in patients with primary pulmonary hypertension, its use in patients with pulmonary hypertension (PHT) and associated congenital heart defects (CHD) has not been evaluated.
Methods and Results—Twenty patients (15±14 years) with PHT and associated CHD (9 atrial septal defect, 7 ventricular septal defect, 4 transposition of the great vessels, 3 patient ductus arteriosus, 3 partial anomalous pulmonary venous return, and 1 aortopulmonary window) who failed conventional therapy (including digitalis; diuretics; oxygen; warfarin; calcium channel blockade, if indicated; and surgery, if operable) were treated with chronic PGI2. Eleven patients had previous cardiac surgery at a median age of 3 years (range, 5 days to 47 years). Eleven of 20 patients had residual systemic to pulmonary shunts. Hemodynamics, NYHA functional class, and exercise capacity were measured at baseline and after 1 year of PGI2 therapy. None of the patients acutely responded to PGI2 administration. Despite lack of an acute response, mean pulmonary artery pressure decreased 21% on chronic PGI2: 77±20 to 61±15 mm Hg (P<0.01, n=16). Cardiac index and pulmonary vascular resistance also improved on long-term PGI2: 3.5±2.0 to 5.9±2.7 L · min−1 · m−2 (P<0.01, n=16), and 25±13 to 12±7 U · m2 (P<0.01, n=16), respectively. NYHA functional class improved from 3.2±0.7 to 2.0±0.9 (P<0.0001, n=19). Exercise capacity increased from 408±149 to 460±99 m (P=0.13, n=14) on long-term PGI2.
Conclusions—Chronic PGI2 improves hemodynamics and quality of life in patients with PHT and associated CHD who fail conventional therapy. As previously demonstrated in patients with primary pulmonary hypertension, long-term PGI2 may have an important role in the treatment of patients with PHT and associated CHD.
Prostacyclin (PGI2), is a potent short-acting vasodilator and inhibitor of platelet aggregation produced by the vascular endothelium. PGI2 decreases pulmonary vascular resistance and increases cardiac output and systemic oxygen delivery when acutely administered to patients with primary pulmonary hypertension (PPH), and this response has been used to determine whether long-term oral vasodilator therapy is warranted.1 2 With the availability of PGI2 for continuous intravenous administration, long-term PGI2 has been shown to improve hemodynamics and quality of life as well as increase survival in patients with severe PPH.3 4 5 6
In 1985, Bush et al first reported using intravenous PGI2 for acute pulmonary vasodilator testing in children with congenital heart defects (CHD) and secondary pulmonary hypertension.7 8 Although he suggested that PGI2 might have a role in the treatment of pulmonary hypertension (PHT) and associated CHD, long-term PGI2 therapy has not been reported for these patients. We are presenting 20 patients with PHT and associated CHD; these patients, having failed conventional therapy, were treated with long-term PGI2.
Twenty patients (15±14 years) with PHT associated with CHD were started on continuous intravenous PGI2 between 1992 and 1996. Both the acute vasodilator testing with PGI2 and long-term PGI2 therapy are Institutional Review Board–approved studies at Columbia University, College of Physicians and Surgeons. Written informed consent was obtained from all adult patients or from parents of each child. In addition to PHT associated with a CHD, patient selection criteria included (1) clinically symptomatic pulmonary arterial hypertension with no evidence of systemic ventricular dysfunction or pulmonary venous hypertension, (2) lack of an acute response with PGI2 testing (as defined below),1 and (3) failure to improve clinically and hemodynamically on conventional therapy. Conventional therapy included digitalis; diuretics; supplemental oxygen; warfarin;9 calcium channel blockade, if indicated;2 and surgical correction, if operable. Untoward effects which precluded the use of calcium channel blockers included right heart failure during acute testing or intolerable symptoms during chronic administration such as nausea, vomiting, dizziness, or headaches. Patients were considered responders to acute PGI2 vasodilator testing if they met the following criteria: (1) ≥20% decrease in mean pulmonary artery pressure (PAPm), (2) increase in cardiac index, and (3) no change or decrease in the pulmonary vascular resistance (PVR) to systemic vascular resistance ratio.1 Previous studies have demonstrated that PPH patients who respond to acute vasodilator testing usually respond to conventional therapy, and patients who are acute nonresponders do not improve with conventional therapy alone.2
Clinical, Exercise, and Hemodynamic Parameters
Clinical assessment (NYHA functional class), exercise capacity (6-minute walk), and hemodynamic parameters were measured at baseline, ie, at the time of initiation of long-term PGI2 therapy; these were repeated after 1 year on continuous PGI2.
Patients underwent right heart cardiac catheterization using standard techniques. Cardiac output was calculated by Fick method with measured oxygen consumption. Before starting long-term PGI2, acute PGI2 testing was performed.1 Venous access for the delivery of long-term PGI2 was obtained by the insertion of a permanent catheter into the subclavian or jugular vein before hospital discharge.3 PGI2 was infused continuously with the use of a portable infusion pump.3 Initial PGI2 dose ranged from 2 to 14 ng · kg−1 · min−1 (mean dose 4±3 ng · kg−1 · min−1, n=20) and was progressively increased to maintain an optimal therapeutic dose similar to its use in patients with PPH.4 After initiation, the dose was subsequently increased until the patient reached a clinical plateau, usually occurring within the first 6 to 12 months after starting PGI2. Thereafter, the PGI2 dose was increased to prevent the recurrence of pulmonary hypertension symptoms, eg, dyspnea on exertion or exercise intolerance. The patients continued conventional therapy, including oral calcium channel blockers (if they had been receiving them before starting PGI2). Sixteen patients underwent repeat right heart cardiac catheterization after 1 year on PGI2.
In 16 patients, hemodynamic data were available either before or during conventional therapy alone. Fourteen of the 16 patients were evaluated at other institutions before referral. These data were used to compare hemodynamics on conventional therapy with baseline hemodynamics obtained just before initiation of PGI2, to determine whether there was hemodynamic improvement with conventional therapy alone. In addition, there were 7 patients who had undergone surgical correction of their CHD for whom preoperative and postoperative data were available.
Data are shown as mean±SD. Statistical comparisons for paired data were analyzed by Student’s paired t test. ANOVA was performed for repeated measures. Post hoc tests of individual differences were performed using Tukey’s procedure. Logistic regression was used to analyze outcome data. Statistical significance was defined as P<0.05.
Baseline clinical, demographic, and hemodynamic parameters are shown in Table 1⇓ (see Appendix⇓⇓ for individual patient profiles). Patients ranged in age from 22 months to 51 years with a mean age of 15±14 years (median age 10 years). There were 8 males and 12 females. A variety of congenital heart defects were present: atrial septal defect (n=9), ventricular septal defect (n=7), transposition of the great vessels (n=4), patent ductus arteriosus (n=3), partial anomalous pulmonary venous return (n=3), and aortopulmonary window (n=1). Six patients had > 1 CHD. Eleven patients underwent surgical repair of their CHD before starting long-term PGI2 at a median age of 3 years (range, 5 days to 47 years). Four patients were <1 year at the time of operation and had not undergone preoperative cardiac catheterization. The remaining 7 patients had preoperative and postoperative hemodynamic studies from other institutions (Table 2⇓). For these 7 patients, the preoperative PAPm was 60±17 mm Hg, systemic arterial pressure mean 86±11 mm Hg, pulmonary vascular resistance 14±6 U · m2, mixed venous oxygen saturation 65±5%, and systemic arterial saturation 92±6%. There were no significant changes in any of these parameters at postoperative cardiac catheterization (2.2±1.6 years; range, 1 to 5 years postoperatively). Surgical repairs for all patients included atrial septal defect (n=4), patient ductus arteriosus (n=2), partial anomalous pulmonary venous return (n=2), ventricular septal defect (n=1), aortopulmonary window repairs (n=1), and arterial switch (n=2), Mustard (n=1), and Senning (n=1) operations. A total of 11 patients had unoperated (n=8) or residual postoperative (n=3) systemic to pulmonary shunts. Of the 9 patients who did not undergo surgical repair, 1 patient had spontaneous closure of a ventricular septal defect by 2 years. The net direction of the shunt at baseline is shown by the Qp:Qs ratios in Table 1⇓. Ten patients were receiving chronic oral calcium channel blockers at the time of initiation of PGI2 (see Appendix⇓⇓). None of the 20 patients had a history of failure to thrive or congestive heart failure in infancy suggestive of a large left to right shunt (eg, significant pulmonary vascular overcirculation, early in life).
Hemodynamic Effects of Acute and Chronic PGI2
Sixteen patients had hemodynamic data available before initiation of conventional therapy or on conventional therapy alone. Serial studies before and during conventional therapy demonstrated no hemodynamic improvement on conventional therapy alone, eg, PAPm increased from 67±22 to 77±20 mm Hg (n=16) on conventional therapy alone.
Sixteen patients underwent repeat right heart catheterization after 1 year on continuous PGI2 (follow up: 15±6 months; range, 8 to 28 months). Three patients were not restudied: poor venous access in 2 and 1 patient declined. One patient died after 4 months while awaiting transplantation (follow-up catheterization at 3 months; follow-up data are not included in the comparative data). The mean dose of PGI2 at follow-up was 82±37 ng · kg−1 · min−1 (n=16; range, 49 to 195 ng · kg−1 · min−1). Hemodynamic effects of acute and long-term PGI2 for the 16 patients who underwent repeat study are shown in Table 3⇓. All patients were nonresponders to acute vasodilator testing and therefore would not be expected to benefit from chronic calcium channel blockers.2 Although there was no significant change in PAPm with acute PGI2 testing (77±20 versus 77±20 mm Hg; n=16), on long-term PGI2 therapy, PAPm decreased 21% (77±20 to 61±15 mm Hg; P<0.01, n=16). In addition, cardiac index increased 69% (3.5±1.6 to 5.9±2.7 L · min−1 · m−2; P<0.01, n=16), PVR decreased 52% (25±13 to 12±7 U · m2; P<0.01, n=16), and mixed venous oxygen saturation increased compared with baseline: (64±7% to 70±8%; P<0.01, n=16). These data illustrate that, similar to PPH patients, lack of an acute PGI2 response does not preclude significant hemodynamic improvement on long-term PGI2.
Two patients (patients 10 and 16) had significant right heart failure, eg, mean right atrial pressures (RAPm) were 15 and 19 mm Hg, respectively, at baseline. On PGI2 therapy, RAPm did not decrease in either: 1 patient died after 4 months. RAPm was 19 mm Hg at repeat catheterization after 3 months, and at repeat catheterization after 1 year in the other patient, RAPm was 17 mm Hg. There were no significant complications in the latter patient, and despite lack of significant improvement in RAPm, all other hemodynamic variables improved. In addition, the patient’s 6-minute walk test (72 to 288 m) and NYHA functional class (IV to II) improved. Overall, RAPm did not change significantly on long-term PGI2 (6±5 versus 8±4 mm Hg; P=0.09, n=16).
Mean systemic arterial pressure did not change on long-term PGI2 therapy (79±12 versus 74±9 mm Hg; n=16, P=NS). In addition, systemic arterial oxygen saturation remained unchanged for the group (91±5% versus 91±7%; n=16), including those patients with residual systemic to pulmonary shunts (90±6% versus 90±7%; n=10). Systemic oxygen delivery improved on long-term PGI2 (580±222 to 940±416 mL · min−1 · m−2; n=16, P<0.001), including those patients with residual shunts (646±185 to 938±409 mL · min−1 · m−2; n=10, P<0.05).
NYHA Functional Class and Exercise Capacity
Seventeen patients were NYHA functional class III (n=10) or IV (n=7) at the initiation of continuous PGI2; 3 were NYHA functional class II. NYHA functional class improved on long-term PGI2: 3.2±0.7 to 2.0±0.9 (n=19, P<0.0001) consistent with increased systemic oxygen delivery. Functional class improved in 14 and remained unchanged in 5 patients. In the one patient who died after 4 months, there was no change in NYHA functional class at 3-month follow-up.
Exercise capacity was assessed using the 6-minute walk test in all patients old enough to perform the test reliably (n=15), ie, patients 5 years or older. Fourteen had repeat testing after 1 year. The 6-minute walk test increased from 408±149 to 460±99 m (n=14; P=0.13).
Complications attributable to PGI2 were frequent and included jaw pain (n=14), rash (n=8), arthralgias (n=6), and nausea/vomiting (n=2). These side effects occurred with similar incidence to PPH patients treated with long-term PGI2.2 3 4 During the first year on PGI2, complications related to the PGI2 delivery system included dislodged central venous lines (n=7), local central venous line infections (n=4), and pump malfunction (n=2). There were no episodes of catheter-related sepsis.
Sixteen patients remain on long-term PGI2 (follow-up 16 months to 5.5 years), 3 patients underwent lung transplantation (after 10, 11, and 17 months, respectively, on PGI2), and 1 patient died after 4 months awaiting transplantation. Eight of the 12 patients listed for transplantation have been taken off the active transplant list because of persistent clinical and hemodynamic improvement. There were no differences in baseline hemodynamic parameters between the patients taken off the active list compared with the patients who remained on the transplant list or died, although this may be due to a small sample size. In addition, one patient with an atrial septal defect (patient 13, see Appendix⇓⇓) went from having an inoperable atrial septal defect to an operable CHD on PGI2 therapy. Her Qp:Qs ratio increased from 2.8:1 to 4.1:1 on long-term PGI2. This patient subsequently had her atrial septal defect partially closed and remains on PGI2. Another institution might have considered closure of the atrial septal defect, given this patient’s baseline hemodynamics. However, on the basis of the overall risk-benefit considerations, the degree of pulmonary vascular disease, and age of presentation (7), we chose to treat with PGI2 with the aim of making the patient operable.
On the basis of previously demonstrated efficacy of long-term intravenous PGI2 therapy in patients with PPH who fail conventional therapy, we present our experience in 20 patients with PHT and associated CHD with pulmonary vascular changes similar to PPH patients.10 Although it is unclear why these patients developed pulmonary vascular disease when other patients with CHD have reversible or much milder pulmonary vascular disease, long-term PGI2 therapy improves quality of life and hemodynamic parameters in patients with PHT and associated CHD who fail conventional therapy.3 5
Conventional therapy for PPH patients has classically included digitalis; diuretics; supplemental oxygen; warfarin; and calcium channel blockade, if indicated. Previous surgical repair of the CHD was included in our definition of conventional therapy for patients in whom surgical repair was performed. All patients treated with long-term PGI2 had failed conventional therapy, clinically and hemodynamically. Previous studies in PPH patients have demonstrated that chronic vasodilator therapy (ie, with calcium channel blockade) in acute responders to vasodilator testing and with continuous infusion of PGI2 in acute nonresponders, as well as in patients who fail conventional therapy despite an initial acute vasodilator response, is effective in improving symptoms, hemodynamics, and survival.2 3 6 11 Thus, our rationale for using PGI2 in patients with PHT and associated CHD was based on similar histopathology in PPH and PHT secondary to CHD even if the pathogenesis might be different.10 Also, the preoperative hemodynamics demonstrated that in addition to severe PHT, the pulmonary vascular resistance was higher in all cases than we would have expected for these patients, reflecting significant pulmonary vascular disease preoperatively. Furthermore, for those patients who had previous surgery and complete repair of their cardiac defects, their postoperative physiology more closely resembled PPH. We speculated that these patients would respond to PGI2 similarly to the experience with PPH patients who fail conventional therapy. Our data supports this with the patients who had PHT and associated CHD responding to long-term PGI2, not unlike PPH patients (eg, significant hemodynamic improvement on long-term PGI2 despite lack of an acute response to PGI2 administration)5 (Table 3⇑).
The definition of PPH, as described by the NIH Registry on PPH,12 excludes patients with CHD: PHT associated with CHD is thought to result from a different mechanism, eg, in patients with systemic to pulmonary shunts, increased pulmonary blood flow with increased pulmonary artery pressure results in increased shear stress leading to pulmonary vascular obstructive changes and shunt reversal. This phenomenon, of PHT secondary to CHD, first described by Victor Eisenmenger in 1897,13 was later termed the Eisenmenger syndrome.14 Previous publications have shown that in most cases of CHD with systemic to pulmonary shunts, early repair of the cardiac defects, (usually by age 2), prevents irreversible pulmonary vascular obstructive disease,15 although exceptions to this teaching have been reported.16 17 In selecting our patients for long-term PGI2 therapy, we speculated that these patients may have had pulmonary vascular disease that was similar to PPH and triggered or exacerbated by the CHD(s). In the patients who underwent early hemodynamic evaluation, pulmonary blood flow was not significantly increased despite significant PHT, suggesting that some patients with CHD may be more prone to develop pulmonary vascular disease than others, even without large systemic to pulmonary shunts. These patients may represent a subgroup in the spectrum of pulmonary vascular disease that exists between PPH and Eisenmenger syndrome; this raises the question whether uncomplicated CHDs, which are not severe enough to cause pulmonary vascular obstructive disease alone (eg, a small ventricular septal defect), may serve as a trigger for the development of PPH.
Our findings may also have implications for treating patients with the classic Eisenmenger syndrome, ie, in whom shunt reversal has occurred. Long-term PGI2 therapy may slow the progression of the Eisenmenger syndrome and render some Eisenmenger patients operable with partial surgical repair leaving an atrial septostomy. This may ultimately improve cyanosis, oxygen delivery, and quality of life. Physicians have been wary about using vasodilators in patients with Eisenmenger syndrome in the presence of systemic to pulmonary shunts because of the possibility of a greater fall in systemic resistance compared with pulmonary resistance resulting in increased right to left shunting. However, in our patients with residual shunts, we did not see an increase in right to left shunting, ie, systemic vascular resistance did not fall more than pulmonary vascular resistance, nor did we see a fall in systemic arterial blood pressure on long-term PGI2. In the 10 patients with residual shunts, there was a significant improvement in oxygen delivery on long-term PGI2. The finding that systemic arterial oxygen saturation did not improve significantly with treatment may be related to the small sample size (n=10).
One of the limitations of our study was an inability to do survival analysis owing to lack of a control group. Sample size was also a limitation for analyzing changes on long-term PGI2 therapy, eg, whether the observed increase in RAPm in 2 patients with PHT and atrial septal defects is clinically significant warrants further study. One might consider the heterogeneity of age and type of CHD another limitation of this study. We would, however, like to emphasize that our observations in this group of patients with a variety of CHDs and ages responding favorably to long-term PGI2 suggests that PGI2 therapy may be effective in a wide range of patients. Although the exact mechanism(s) of action remains unclear, the indications for long-term PGI2 may be broader than the current recommendations. Studying the unusual patient, as opposed to the typical patient with PPH, may offer insight into why some patients develop pulmonary vascular disease and others do not; it may also offer clues as to how long-term PGI2 may remodel the pulmonary vascular bed.
In conclusion, long-term PGI2 improves hemodynamic and quality of life parameters in patients with PHT and associated CHD who fail conventional therapy. In addition to its use as a palliative bridge to transplantation, long-term PGI2 may also be an alternative to transplantation in selected patients, enabling previously inoperable patients to become operable. Our observations suggest that long-term PGI2 may play an important role in the treatment of patients with PHT and associated CHD, just as it does for PPH patients. We would like to emphasize that although these patients (all of whom were nonresponders to acute vasodilator testing and failed conventional treatment) improved on long-term PGI2, all patients with PHT and associated CHD should not be empirically treated with long-term PGI2. On the basis of overall risk-benefit considerations, patients should first undergo evaluation, including acute vasodilator testing, and be considered for long-term PGI2 only if they are acute nonresponders during vasodilator testing or fail conventional therapy regardless of their acute vasodilator response.
This study was supported in part by the Hatch Young Cardiovascular Medicine Training Fellowship research grant and a grant from the National Institute of Health, National Center for Research Resources: RR-00645.
- Received August 20, 1998.
- Revision received December 1, 1998.
- Accepted December 29, 1998.
- Copyright © 1999 by American Heart Association
Barst RJ, Rubin LJ, Long WA, McGoon MD, Rich S, Badesch DB, Groves BM, Tapson VF, Bourge RC, Brundage BH, Koerner SK, Langleben D, Keller CA, Murali S, Uretsky BF, Clayton LM, Jobsis MM, Blackburn SD Jr, Shortino D, Crow JW. A comparison of continuous intravenous epoprostenol (prostacyclin) with conventional therapy for primary pulmonary hypertension. N Engl J Med. 1996;334:296–301.
Rubin LJ, Mendoza J, Hood M, McGoon M, Barst RJ, Williams WB, Diehl JH, Crow J, Long W. Treatment of primary pulmonary hypertension with continuous intravenous prostacyclin (epoprostenol). Ann Intern Med. 1990;112:485–491.
Bush A, Busst C, Booth K, Knight WB, Shinebourne EA. Does prostacyclin enhance the selective pulmonary vasodilator effect of oxygen in children with congenital heart disease? Circulation. 1986;74:135–144.
Fuster V, Steele PM, Edwards WD, Gersh BJ, McGoon MD, Frye RL. Primary pulmonary hypertension: natural history and the importance of thrombosis. Circulation. 1984;70:580–587.
Rich S, Dantzker DR, Ayres SM, Bergofsky EH, Brundage BH, Detre KM, Fishman AP, Goldring RM, Groves BM, Koerner SK, Levy PC, Reid LM, Vreim CE, Williams GW. Primary pulmonary hypertension: a national prospective study. Ann Intern Med. 1987;107:216–223.
Eisenmenger V. Die Angeborenen defecte der kammerscheidewand des herzen. Z Klin Med Suppl. 1897;32:1–28.
Wood P. The Eisenmenger syndrome or pulmonary hypertension with reversed central shunt. Br Med J. 1958;2:701–709, 755–762.
Blieden LC, Moller JH. Small ventricular septal defect associated with severe pulmonary hypertension. Br Heart J. 1984;52:117–118.